Disruption of a sugar transporter gene cluster in a hyperthermophilic archaeon using a host-marker system based on antibiotic resistance

Abstract

We have developed a gene disruption system in the hyperthermophilic archaeon Thermococcus kodakaraensis using the antibiotic simvastatin and a fusion gene designed to overexpress the 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase gene (hmg(Tk)) with the glutamate dehydrogenase promoter. With this system, we disrupted the T. kodakaraensis amylopullulanase gene (apu(Tk)) or a gene cluster which includes apu(Tk) and genes encoding components of a putative sugar transporter. Disruption plasmids were introduced into wild-type T. kodakaraensis KOD1 cells, and transformants exhibiting resistance to 4 microM simvastatin were isolated. The transformants exhibited growth in the presence of 20 microM simvastatin, and we observed a 30-fold increase in intracellular HMG-CoA reductase activity. The expected gene disruption via double-crossover recombination occurred at the target locus, but we also observed recombination events at the hmg(Tk) locus when the endogenous hmg(Tk) gene was used. This could be avoided by using the corresponding gene from Pyrococcus furiosus (hmg(Pf)) or by linearizing the plasmid prior to transformation. While both gene disruption strains displayed normal growth on amino acids or pyruvate, cells without the sugar transporter genes could not grow on maltooligosaccharides or polysaccharides, indicating that the gene cluster encodes the only sugar transporter involved in the uptake of these compounds. The Deltaapu(Tk) strain could not grow on pullulan and displayed only low levels of growth on amylose, suggesting that Apu(Tk) is a major polysaccharide-degrading enzyme in T. kodakaraensis.

FIG. 1.

The mevalonate pathway for isoprenoid lipid biosynthesis in Archaea. The reaction catalyzed by HMG-CoA reductase is boxed with dotted lines. It should be noted that the two reactions converting mevalonate phosphate to isopentenyl diphosphate are distinct from the reactions in the classical mevalonate pathway (26).

FIG. 2.

Disruption of the apu_Tk and mal_Tk loci of T. kodakaraensis. (A) Design of the hmg_Tk overexpression cassette using the 5′ upstream flanking region of gdh_Tk. (B) Gene organization of the putative maltooligosaccharide transporter of T. kodakaraensis. TK1774* represents the correct sequence of the apu_Tk gene (see text). Black arrows indicate the gene(s) disrupted in this study. (C) The two plasmids constructed for the disruption of the apu_Tk and mal_Tk loci via double-crossover recombination. (D) PCR analyses of the apu_Tk and mal_Tk loci confirming gene disruption. Primers were designed in the 5′ and 3′ flanking regions of the gene(s) to be disrupted. DNA size markers were run in lane M, and their sizes (bp) are indicated to the left of the gels. The results of PCR with wild-type T. kodakaraensis KOD1 and five individual transformants are indicated in lane W and lanes 1 to 5, respectively. The arrowheads to the right of the gels indicate the amplified fragments expected before and after recombination. The decreases in lengths of the amplified fragments reflect the differences in length between apu_Tk (∼3,500 bp) and P_gdh-hmg (∼2,000 bp) and between mal_Tk (∼7,000 bp) and P_gdh-hmg. Nonspecific amplifications of DNA fragments observed for the wild-type mal_Tk locus were due to the prolonged reaction time necessary to amplify the entire locus.

FIG. 3.

Growth of wild-type T. kodakaraensis KOD1 and Δapu_Tk and Δmal_Tk mutant strains in the presence of various concentrations of simvastatin. Open circle, wild-type strain; filled square, Δapu_Tk strain; filled triangle, Δmal_Tk strain; OD, optical density.

FIG. 4.

Growth of wild-type T. kodakaraensis KOD1 and Δapu_Tk and Δmal_Tk mutant strains on various carbon sources. The carbon sources examined are indicated above each panel. Glucose and maltose were not examined, as the wild-type strain cannot utilize these sugars. Open circle, wild-type strain; filled square, Δapu_Tk strain; filled triangle, Δmal_Tk strain; OD, optical density.

FIG. 5.

Southern blot analyses on Δapu_Tk strains obtained with the hmg_Pf gene as a selectable marker. Genomic DNA from five selected Δapu_Tk strains and from wild-type T. kodakaraensis KOD1 (W) was subjected to Southern blot analyses using probes within the regions indicated below each membrane.

FIG. 6.

Carboxy-terminal regions of various amylopullulanase proteins from the Thermococcales, along with the corresponding regions of periplasmic components of two putative ABC-type dipeptide transport systems of T. kodakaraensis. All of these proteins harbor a threonine (or serine)-rich region, followed by a putative transmembrane domain and a stretch of basic residues (indicated by circles) at the extreme C terminus. The subscripts of the amylopullulanase proteins identify the source organism as follows (accession number): Apu_Tl, Thermococcus litoralis (BAC10983); Apu_Th, Thermococcus hydrothermalis (AAD28552); Apu_Pf, P. furiosus (ABA33719); Apu_Pa, P. abyssi (CAB49104). Accession numbers for TK1760 and TK1804 are BAD85949 and BAD85993, respectively.